The commonest gastrointestinal (Gl) carcinoids occur in the small intestine and appendix. Both are derived from the enterochromaffin (EC) cell, but their malignancy and propensity to metastasize vary widely. Small intestinal carcinoids (SICs) are aggressive and exhibit hepatic and lymph node (LN) metastasis;appendiceal carcinoids (APCs) are slow-growing and rarely metastasize. Gastric carcinoids (GCs) [enterochromaffin-like (ECL) cells], are slow-growing and rarely metastasize unless gastrin-independent. No biological explanation for the variable malignancy of Gl carcinoids exists. We hypothesize that a gene expression-based definition of these tumors will: 1) allow prediction of malignancy and metastasis, and 2) provide accurate staging and facilitate rational therapy. Utilizing: i) GeneChip analysis, ii) quantitative real-time PCR (Q RT-PCR), and iii) carcinoid tissue microarray (TMA) immunostaining/quantitation, we evaluated gene and protein expression in Gl carcinoids. In SICs and their metastases, the neuroendocrine marker chromogranin A (CgA);NAP1L1 (mitosis regulator);MAGE-D2 (liver metastatic predictor), and MTA1 (metastasis, migration), are over- expressed. In malignant APCs and GCs, CgA, MAGE-D2, MTA1, and NAP1L1 are all significantly over- expressed. In 30% of histologically-negative lymph nodes, molecular analysis detected CgA transcript and protein, making sensitive molecular staging of carcinoid tumors possible. This proposal seeks, using Q RT- PCR of frozen and paraffin specimens and carcinoid TMA analysis, to define the malignant potential of Gl carcinoids. The aims are: 1) Confirm the specificity of the neuroendocrine cell marker gene, CgA, in identifying Gl carcinoids, 2) Determine whether increased levels of the mitotic and metastasis-associated genes, NAP1L1, MAGE-D2 and MTA1, characterize malignant Gl carcinoids, 3) Determine whether benign and malignant appendiceal carcinoids show predictive differential gene expression, and 4) Identify occult metastasis in histologically normal lymph node or liver biopsies using Q RT-PCR. We propose that delineation and quantification of gene expression will: 1) define malignancy of an individual carcinoid and 2) identify occult metastasis. This will provide novel biological information and provide molecular data to identify occult metastasis predict spread and facilitate preemptive appropriate therapy.